As the demand for high-speed and wirelesstransmission of information is rapidly increasing, spectrally efficient digiral radio modulation and demodulation scheme is preferred. In this paper, we have studied the signal generation, demodulation and performance evaluation of spectrally efficient QAM signals in a wireless channel environment. We propose differential modulation and demodulation scheme for QAM in rayleigh fading channel. First, formulas for theerror probability of proposed QAM scheme are derived over QWGN channel. Second, we evaluate the performance of the proposed QAM scheme in rayleigh fading environment through computer simulation, and confirm that it shows much better performance compared to that of conventional QAM scheme.

In the case of redundancy control performed by a netowrk management system, it can not be ignored some delay of control time and break of control by failure of the system. This paper proposed a new method to improve the performacne of a redundancy control system interfaced to the network management system. At first, it produced the modelling of the total system and analyzed the functional relation of reliability and failure rate for a transmission system having an one-to-one active redundancy structure controlled by itself. After describing the loss rate of transmission data and the time required for the redundancy control process of the network management system when a failure is occurred, it proposed the new redundancy method of complex system control mode as an effective redundancy control method to improve reliability of a redundancy control system. Finally, it proved the effect by comparing the new result to the previous result.

In this paper we propose a new dual structured tap selection algorithm for voice echo canceller in digital cellular communication system, investigating adaptive filtering algorithms for echo cancellation in long distance telephony or mobile communication system. The proposed algorithm has a two-stage processing structure that after a dispersive region in an impulse response of an echo path is found out, the tap coefficients of a short length filter are adjusted adaptively for the region, because the impuse response has a very little portion of the dispersion. Simulation results show that the proposed algorithm with 256 taps gives a performance of convergence speed superior to both full-tap normalized least mean with 256 taps gives a performance of convergence speed superior to both full-tap normalized least mean square (NLMS) and a scrub taps waiting in a queue (STWQ) algorithms by about eighty per cent, also to a tap selection algorithm by about twenty per cent. And the resutls diplay that if the more tap coefficients are used due to a long delayed dispersive zone, the proposed algorithm produces the better performance.

It is well known that the viterbi algorithm proposed as a mthod of decoding convolutional codes is in fact maximum likelihood (ML) and therefore optimal. But, because hardware complexity grows exponentially with the constraint length, there will be severe constraints on the implementation of the viterbi decoders. In this paper, the three-layered backpropagation neural networks are proposed as an alternative in order to get sufficiently useful performance and deal successfully with the problems of the viterbi decoder. This paper shows that the neural convolutional decoder (NCD) can make a decision in the point of ML in decoding and describes simulation results. The cause of the difference between stochastic results and simulation results is discussed, and then thefuture prospect of the NCD is described on the basis of the characteristic of the transfer function.

In this paper, we analyzed the performance improvement of the DS-CDMA/dPSK cellular mobile communication system with CCI canceller and RAKE receiver techniques in mobile radio channel which is characterized by AWGN, Multi-user interference (mUI) and frequency-selective rayleigh fading. System capacity i.e., number of user per cell has been derived and the evaluated results are shown in figuraes as a function of PN code sequence length, number of RAKE receiver tap, BER and E$_{b}$/N$_{o}$. The voice activity factor is assumed to be 3/8 the number of sectors in a cell 3 and MUI is modeled as gaussian process. From the results, the capacity of the DS-CDMA/DPSK cellular mobile communication system is improved by adopting CCI canceller. When we adopt the RAKE receiver to mitigate the multipath fading, the error performance is improved and the amount of improvement is proportional ot the number of taps L in the case that each tap contains a signal component and the better of improvement can be obtained, the smaller of the SNR difference in each tap. The system capacity is more increased when the CCI canceller and the RAKE receiver is used in cascade.

For the implementation of callcontrol software, it will be shown how to divide problems into dimain and also presents methodologies to specify domains according to the object-orented analysis techniques. Domains is specified by the set of related objects as shown in call control examples. Events and conditons are able to be transformed into system invariant so that is possible to figure out the relationship between domains and those concludes that the software requirements specification can be established for the design and implementation. Relationships existing between domains can be formalized so as to verify software requirements.

This paper aims to support delay-sensitive multimedia applications by suggesting a mechanism in which maintains almost constant end-to-end delay thus providing the optimum playout synchronization. For this task the sum of network delay and buffering delay is entiredly managed and to eliminate little delay fluctuations and instantaneous delays at a buffer and a network, a low pass filter is used. Furthermore the correction function, which is used for maintaining the buffering level ot a reference value, is a non-linear step function, unlike the existing linear and continuous function. it has a different step sizes adapting to a traffic characteristics of a network congestion. the proposed mechanism has been confirmed of it sefficiency through SLAM-II netowrk.

In this paper, we investigate the effect of a LAN traffic on the performance of a shared buffer ATM switch andIWF (interworking function )on a LAN environment through simulations. Firstly, the delay and the buffer occupancy of the switch and IWF are mesured according to the proportion of the LAN traffic to the traffic generated by gernoulli process. Secondly, we investigate the behavior of the switch in the case that LAN traffic is concentrated to a connectionless server, and the effect of LAN traffic shaping at IWF on the delay and the buffer occupancy of the switch.

A timing recovery circuit of 10 Gbit/s optical receiver is described. The circuit consists of a passive NRZ-to-PRZ circuit, a dielectric resonator filter (DRF) and a narrow band amplifier, which for the first time adopted a temperature compensation technique using the tempareature characteristics of DR. The experimental results showed an output clock phase variation of less than ${\pm}$6 degree over the operating temperature range form 0$^{\circ}C$ to 75$^{\circ}C$ and measured maximum rms jitters of less than 2 phs with the resonance detunings of up to ${\pm}$10 MHz. These experimental results show that the circuit is a suitable for 10 Gbit/s lightwave transmission system.

Input filter for satellite communication transponder is the interface between the antenna and the receiver. It is used to provide the selection of the uplink signals with minimum insertion loss and to prevent downlink signals form reaching the LNA. This paper is intended to provide a description of the input filter for KOREASAT communication transponder. Included are description for the electrical and mechanical design and the requirments of environmental test. In expecting the electrical performine the optimum electrical configuration ot meet all requirements are performed. Mechanical requirements are charactersed by several constraints for weight, size of the filter and its type of input output interface. The standardized environmental tests are performed to confirm satisfactory performance of the filter with respect to the requirements of vibration and thermal vacuum shocks.

The attenuation pattern of the E-polarized electromagnetic fields scattered by two empty circular cylinders in a dielectric medium is analyzed to calculate the effects of the inhomogenuity in hte background medium. Inthe amplitude pattern of the electromagnetic field scattered by one cylinder, double dips occur at the locations corresponding to its top and bottom boundaries. The distorton of the original double dip pattern is calculated when other air cavity approaches or its size increases.

The design method of 85 GHz-115 GHz band single side-band filter using the principle of martin-puplett inteferometer is described. It has been designed by this method and also manufactured and tested. From the test results, not only the ratio of image singal rejection of 19 dB is obtained, but also the theoretical and experimental results of center frequency of pass-band and rejection-band show the validity of the theory. This manufactured filter was installed on 100GHz band SiS (superconductor insulator superconductor ) receiver for observing cosmic radio waves and tested. We found that this filter can be used a single side-band as well as double side-band mode. The design method which is presented in this paper can be used a single side-band filter for a heterodyne type sub-milimeter wave receiver.

Schottky diodes are fabricated on laser annealed and unannealed polysilicon substrate and their electrical characteristics are studied and analyzed. Current of laser annealed devices are larger than that of unannealed devices because of grain growth, decrease of grain boundary and trap density, lowering of grain boundary barrier height, decrease of dopant segregation. At low forward bias (<0.7V), currents of unanealed devices are larger. Soft breakdown voltages of laser annealed devices are larger than that of unannealed devices.

With the MOsES (mask oxide sidewall etch scheme)process which uses the conventional i-line stepper and isotropic wet etching, CMOSFET's with fine gate pattern of 0.1.mu.m CMOSFET device, the screening oxide is deposited before the low energy ion implantation for source/drain extensions and two step sidewall scheme is adopted. Through the characterization of 0.1.mu.m CMOSFET device, it is found that the screening oxide deposition sheme has larger capability of suppressing the short channel effects than two step sidewall schem. In cse of 200.angs.-thick screening oxide deposition, both NMOSFET and PMOSFET maintain good subthreshold characteristics down to 0.1.mu.m effective channel lengths, and show affordable drain saturation current reduction and low impact ionization rates.

The impact ionization rate of thethighly-doped AlGaAs/GaAs quantum well structure is calculated, which is an important parameter ot design theinfrared detector APD and the novel neural device. In conjunction with ensemble monte carlo method and quantum mechanical treatment, we analyze the effects of the parameters of quantum well structure on the impact ionization rate. Since the number of the occupied subbands increases while the energy of the subbands decreases as the width of quantum well increases, the impact ionization rate increases in the range of th esmall well width but gradually the increament slows down and is finally saturated. Due to the effect of the energy of the injected electrons into the quantum well and the tunneling through the barrier, the impact ionization rate increases for the range of the small barrier width and decreases for the range of the large barrier width. Thus, there exists a barrier width to maximize the impact ionzation rate for a mole fraction x, and the barrier width moves to the larger vaue as the mole fraction x increases. The impact ionization rate is much more sensitive to the variation of the doping density than that of the other quantum well parameters. We found that there is a limit of the doping density to confine the electronics in the quantum well effectively.

In this paper, the hot carrier effect and device degradation of deep submicrometer SC-PMOSFETs have been measured and characterized. It has been shown that the substrate current of a 0.15$\mu$m PMOSFET increases with increasing of impact ionization rate, and the impact ionization rate is a function of the gate length and gate bias voltage. Correlation between gate current and substrate current is investigated within the general framework of the lucky-electron. It is found that the impact ionization rate increases, but the device degradation is not serious with decreasing effective channel length. SCIHE is suggested as the possible phusical mechanism for enhanced impact ionization rate and gate current reduction. Considering the hot carrier induced device degradation, it has been found that the maximum supply voltage is about -2.6V for 0.15$\mu$m PMOSFET.

In this paper, a new CMOS on-chip current reference circit for memory, operational amplifiers, comparators, and data converters is proposed. The reference current is almost independent of temeprature and power-supply variations. In the proposed circuit, the current component with a positive temeprature coefficient cancels that with a negative temperature coefficient each other. While conventional curretn and voltage reference circuits require BiCMOS or bipolar process, the presented circuit can be integrated on a single chip with other digiral and analog circits using a standard CMOS process and an extra mask is not needed. The prototype is fabricated employing th esamsung 1.0um p-well double-poly double-metal CMOS process and the chip area is 300um${\times}$135 um. The proposed reference current circuit shows the temperature coefficient of 380 ppm/.deg. C with the temperature changes form 30$^{\circ}C$ to 80$^{\circ}C$, and the output variation of $\pm$ 1.4% with the supply voltage changes from 4.5 V to 5.5 V.

This paper presents a new symmetrical routing architecture for FPGA and an efficient routing algorithm for the architecture. The routing architecture adopts the segmented wires and the improved switch modules. Segmetned wires construct routing channels which pass through the chip in vertical and horizontal directions. To maximize the utility of a track, a track in each switch module can be separated in two part using a programmable switch to route two different net. The proposed routing algorithm finds all assignable tracks for a given net and selects the best track from assignable tracks to minimize the number of programmable switches and the unused portion of the wire segments. In order to stabilize the perfomrance of the algorithm, the routing order is defined by weighted sum of the number of wire segment, the length of wire segmetn, and the number of pin. Experimental results show that routability is improved dramatically and the number of crossing switches are reduced about 40% compared with the previous works.

In tis work, a three-stage pipelined A/D converter (ADC) was implemented to obtain 10-bit resolution at a conversion rate of 20 msamples/s for video applications. The ADC consists of three identical stages employing a mid-rise coding technique. The interstage errors such as offsets and clock feedthrough are digitally corrected in digitral logic by one overlapped bit between stages. The proposed ADC is optimized by adopting a unit-capacitor array architecture in the MDAC to improve the differential nonlinearity and the yield. Reduced power dissipation has been achieve dby using low-power latched comparators. The prototype was fabricated in a 0.8$\mu$m p-well CMOS technology. The ADC dissipates 160 mW at a 20 MHz clock rate with a 5 V single supply voltage and occupies a die area of 7 mm$^{2}$(2.7 mm $\times$ 2.6mm) including bonding pads and stand-alone internal bias circuit. The typical differential and integral nonlinarities of the prototype are less than $\pm$ 0.6 LSB and $\pm$ 1 LSB, respectively.